12
$\begingroup$

I'm at a critical point in a sci-fi novel where a couple of thousand Superintelligent artificial intelligences end up going haywire. Naturally, they're extremely intelligent, and thus, much harder to deal with than an average human.

Up to here, these AIs have been almost docile. The power lust of the man who created them forces these AIs into servitude, and then he leverages them like an army against an attacking force in order to defend a city.

This city is effectively an experiment put into action years ago by six scientists, who lead the entire place. Since the whole city counts as a private place, run by a private company, the US government has only some minor influence inside.

The nebulous part is this attacking force. Even if I was from the US, no US defense organization would really list 'to stop invasions, insurrections, rebellions and other forms of conflict caused by goddamn robots'. My best guess is Homeland Security, which sounds suited to responding to internal threats like this. What do you think?

$\endgroup$
14
  • 1
    $\begingroup$ What is specifically attacked? Depending on the target, it could be FBI or Secret Service or local police that would respond to it. Or is your question "Which US agency is best equipped to defend against a hacking invasion?" $\endgroup$
    – Alexander
    Commented May 3, 2019 at 21:51
  • 1
    $\begingroup$ Is the AI's creator one of the six scientists? $\endgroup$
    – Alexis
    Commented May 3, 2019 at 21:53
  • 1
    $\begingroup$ I find it really not clear who's attacking who? Are the AI attacking the city or defending it? $\endgroup$
    – Alexis
    Commented May 3, 2019 at 21:55
  • 4
    $\begingroup$ Unless the government in your world is radically different, the US doesn't have "only some minor influence" in its own cities. Even if they're organized as private entities, they're still bound by state and federal law. $\endgroup$
    – Cadence
    Commented May 3, 2019 at 22:03
  • 3
    $\begingroup$ The AIs have been leveraged by one of the six scientists in order to defend the city. Boiling down the US's objective, they really just want to bring the AIs under control, with the intent of destroying all of them and then taking control of the city. The plan to actually do that is a pretty long-winded one, but to sum it up, once all AIs are either captured or destroyed, the US wins. The scientist and AI slaver is trying to keep that from happening. $\endgroup$
    – Grizzly
    Commented May 3, 2019 at 22:04

8 Answers 8

21
$\begingroup$

Invent it.

Big changes have taken place in the US, I gather. If there are thousands of superintelligent AIs milling about to be conscripted, over previous years there were probably other issues involving AIs and semisentitent automatons.

You can invent your government agency to deal with these. Heck, The Department of Motherland Security was only invented after 911 so the US is allowed to make new ones. Then you will also have some immunity from any concrete-thinking readers who protest "hey! hey! The NSA is not allowed to carry stun guns!" because you have invented your agency and its rules for your near future scenario. I suspect this new agency would itself employ some Daneel Olivaw type AIs as well as regular detectives. The agency personnel sometimes bump up against the older branches like the FBI, with political shenanigans, prejudice and backbiting helping to leaven your story.

Maybe you could repurpose ATF for your agency, so you could have some continuity with existing institutions. They could be Alcohol Tobacco Firearms and Robots.

$\endgroup$
5
  • 4
    $\begingroup$ Just to add to this, it is also plausible that, in a world where there is a high risk of cyber attacks, it is likely an organisation or department would already have been created to deal with these attacks long before Ai was developed. This pre-existing organisation, originally created to deal with any cyber attacks, would also deal with attacks from Ai. There may even be a branch within said organisation that specifically deals with rouge Ai attacks. $\endgroup$ Commented May 3, 2019 at 22:34
  • $\begingroup$ Very good point. There's definitely been enough time and plenty of AI-related stimuli to force the gov into making an agency like this. $\endgroup$
    – Grizzly
    Commented May 3, 2019 at 22:36
  • 2
    $\begingroup$ I upped this answer simply because a personal annoyance of mine is when authors misappropriate existing government agencies because they're too (lazy, unimaginative, etc) to create something more interesting. It's certainly plausible that the US Dept of Education may have an alien task force for some reason that only makes sense to bureaucrats in D.C. but is that compelling? Make up your own dark government entities and see where that takes you. It's a bit cliche perhaps but at least it's not ridiculous. $\endgroup$
    – dhinson919
    Commented May 4, 2019 at 16:54
  • $\begingroup$ @dhinson919, in this case, misappropriation isn't a problem because there are existing agencies to deal with insurrection. First line of response is the appropriate police agency (depending on the details, probably either the county sheriff's department, the state police, or the FBI). Once they're unable to handle it, it escalates to the state National Guard, and from there, to the US Army. $\endgroup$
    – Mark
    Commented May 4, 2019 at 18:42
  • 3
    $\begingroup$ Stargate SG-1/Atlantis gave us the Department of Homeworld Security to cover this issue; i.e; good, professional grade advice right here: Invent it. $\endgroup$ Commented May 4, 2019 at 21:20
8
$\begingroup$

The first people on the scene will almost certainly be law enforcement of some kind. Accordingly, the people who are sent in will depend on the laws they think the science team has broken.

The act of setting up a new, company-owned city in the middle of nowhere, as opposed to establishing oneself in an existing city (which would be, by all accounts, far cheaper and easier) is going to raise a lot of red flags. Probably the most benign case is that they're using it to launder money and/or dodge taxes on a grand scale. That will get the FBI (in its anti-racketeering role) and the IRS interested in their operations.

Once they're actually accused of a federal crime and a warrant issued, it becomes the job of the US Marshals Service to bring in the suspect(s) to stand trial.

On the other hand, if at any point the scientists want to equip their robot army with anything more menacing than various semiautomatics, they're going to end up breaking any number of federal gun laws. (It's also possible that, seeing the potential threat they pose, the government declares that the robots themselves are restricted munitions.) This will get the ATF involved right quick, and that will tend to bring events to a head fairly soon - gun crimes being both dangerous to the federal agents and not that complicated to prove.

If law enforcement feels that it needs more firepower, it can call in the National Guard (including state National Guards) for backup. Under the Posse Comitatus Act, actual-as-such armed forces - the Army, Air Force, Navy, and Marines - are not allowed to act in a law enforcement role within the US. However, they can act in support and provide information and logistics, so e.g. the FBI could ask for drone surveillance of the city, but could not request troops to be on hand when making arrests.

However, this is only a temporary condition. Under the Insurrection Act, the military can be deployed in a condition of "insurrection, domestic [i.e., internal, not international] violence, unlawful combination, or conspiracy" if the state proves unable to do so on its own. There's red tape involved, and the President (who makes the call in this situation) might want to request an act of Congress for the show of things. However, if it proves absolutely necessary they can be deployed.

(There's also an exemption in the law for cases involving nuclear materials. Hopefully your robots aren't fission-powered.)

Stepping from worldbuilding to writing for a moment, my personal feeling would be to put the ATF in the drivers' seat. Drawing a comparison to Ruby Ridge or Waco, writ however large, makes the government response sound more plausible.

$\endgroup$
1
  • $\begingroup$ Adding to an excellent answer: Depending on which side you intend the villain to be, you could also read about The Battle of Blair Mountain. I personally had no idea such things had happened in United States history until recently. $\endgroup$
    – UrQuan3
    Commented May 7, 2019 at 20:13
4
$\begingroup$

Since this is an experiment with expensive equipment and intellectual property, your company will probably need to hire a significant security force of their own. This is to prevent industrial espionage.

With the amount of area (city sized) that you're talking about, you'll need a lot of these people. As many cameras as you are likely to have, you'll still need/want roving guards to keep an eye out for anything suspicious.

With this security force available, you may want to have a weapons locker. It's not that you think you'll ever need it, but it's just for emergencies.

With these people and weapons available, once you get attacked, you respond with force. If it gets out of control, the police are next to respond. Beyond that is the National Guard. If the local unit isn't enough, they will pull in more units. If it gets further than that, it's a Serious Problem that likely has become a problem for more than just you and needs National help. The Marines are the first to respond to that, then the Army, then Air Force.

Of course, your company will likely have to reimburse the branches for their efforts, so you'd better hope that your hired security can handle things.

$\endgroup$
2
  • $\begingroup$ As many cameras as you are likely to have, any source to back this up? This might not be the smartest move when your project infrastructure may come into contact with outside networks (and with people on-site it likely will be, even if by accident). $\endgroup$
    – JJJ
    Commented May 4, 2019 at 22:18
  • 1
    $\begingroup$ @JJJ, that's more from personal experience and observation than anything. Prisons, hospitals, banks, grocery stores, and more have all kinds of cameras, but many also have security personnel wandering around. I know it sounds like a movies to suggest that people trying to do bad things avoid, disable, etc. cameras, but where do you think Hollywood got that idea? If this is a major installation with high tech breakthroughs, there's going to be industrial espionage. That can hard to see on a camera, but if you can be face to face with employees and see them nervous, that's more obvious. $\endgroup$ Commented May 6, 2019 at 5:29
3
$\begingroup$

The National Guard is the highest level of response I'd expect to show up unless things have really gone south. The Army/Navy/Airforce are all specifically prohibited against operating on American soil, and things need to be at Civil War or invasion levels before those prohibitions get lifted. Basically, you only get the Army roll into town if the incident is a threat to the US as a whole.

Assuming he only wants to control the city area, this will likely be treated as if someone barricaded their house and took hostages, but on a larger scale. Objectives will be to remove everyone possible from the area, and talk the guy down. (While our scientists attempt to break his control over the AIs.) Objectives will get a lot more bloody a lot faster if he's expanding the area he controls, or is killing people, but without that aspect, the goal will be to keep things peaceful.

$\endgroup$
1
  • 1
    $\begingroup$ We call the Civil war/invasion an Insurrection and, yes, the full military (and you missed the Marines) could get involved then. $\endgroup$ Commented May 3, 2019 at 22:46
2
$\begingroup$

Every branch of the military has an engagement plan. They're already looking in to how to deal with a space threat that hits the US, today. They're constantly searching for new ways they can be attacked, and they have engagement specialists looking in to it. every kind of battlefield, every kind of engagement.

Regardless of who is actually engaging in the fighting, they can all talk. National guard may be on the front lines, but they can bring in a specialist from the marines who is in the command room giving advice on how they've simulated this before.

You can essentially do whatever you want. And if you want to have the airforce come in, call an emergency session of congress. Why would they have an emergency session at congress? because the CIA has already been monitoring this threat, and they're putting their response into action.

By the way, every branch of the military can own a war machine. Planes aren't limited to the airforce and ships aren't limited to the navy.

To win over your military audience. Consider engaging in their own language. Watch a few interviews with military persons.

Look at this amazing interview with a 4-star general

$\endgroup$
2
$\begingroup$

Similar to Civil War II situation: Presidential powers to declare national Emergency and Martial Law would handle putting all government forces on case.

But realistically no human could fight against an AI or worse cabal of AIs that was seriously attacking without constraints. INSTANT ARMAGEDDON!!!

All targets would be fired upon as fast as mechanically possible with totally accurate targeting, de-confliction (no interference between attacks) and prioritization of targets (fastest reacting and highest value stuff hit first).

Humans just could not compete with execution and reflexes in nanoseconds and targeting that starts at best existing DOD weapon systems -- then improves on it. Once AI makes decision (the long part of process) the execution has no delays. Then add that human society would be hit by total surprise as there is no human personal activities, emotions, and body language to read. Plus people would assume any computer actions were authorized by some human.

So as always your AI war should have AI motives behind it beyond simple destruction or totally ruthless power grab without limits. That is the AI want most humans to survive and are manipulating people to do or act in a certain way. Thus the AIs are mostly waiting on humans to make next chess move. Oh and maybe the AIs are burning a lot of cycles figuring out what humans are thinking or how humans will react to possible AI moves.

So a good AI war story is more about PSYCHOLOGICAL WARFARE than the physical fight for survival that the AIs can win almost every time at least in the short run. Its more about what AIs and humans think the other side is willing to give up to to reach an accord. Are humans willing for most humans to be slaves? How much territory will AIs set aside as a human reservation? At least that is my opinion.

$\endgroup$
2
  • 4
    $\begingroup$ AI is not magic. AI must obtain Ammo, Fuel and Weapons same as humans. Takes lots of time and industry. Selecting targets may be instant (Pentagon maintains pre-compiled lists), hitting target is limited by speed of the chosen weapon and distance to target, same as for people. Inventing a new weapon needs prototyping and testing. Simulated results do not match reality. $\endgroup$ Commented May 4, 2019 at 0:21
  • 1
    $\begingroup$ @Magicianeer, AI still has way more power in purely informational areas. It can control communications and can fake any message. They may not even need a single round of ammo, humans army will be shooting itself. Imagine fake messages from president, Pentagon, a lot of fake news across the country, banks hacked, war machines purchased from private shops... Heck, military jets are computer controlled (bombs out!). I'm not even talking about releasing chlorine into the atmosphere from automatic factories. No, meatbags, the next step of evolution will only take days. $\endgroup$ Commented May 4, 2019 at 0:44
1
$\begingroup$

The local building inspector and zoning board. Or maybe the angry neighbor next door.

You cannot build an enormous robot city without local regulators getting involved. This is why very few supervillains operate within the USA anymore, and only ordinary villains remain.

...

"Hi, building inspector? This is Greg over at the Power Company. We've got a connection request over on Mongoose Road for some kind of big factory or industrial site, but they don't seem to have a building permit from you guys."

...

"Look, Doctor Crazy, I don't care what kind of super-robots you want to build to subjugate humanity. You need eye-wash stations in the labs, you need clearly labeled fire extinguishers in the hallways, you need exit doors that open outward, and you need to provide me as-built plans. Or next time I come by there's gonna be a fine involved."

...

"Gee, Doctor C., you built this industrial compound illegally on land zoned for agricultural use. So the Tax Appeals Board is not going to consider your request for a reduced tax assessment. And threatening the Board doesn't help - that just gets the State Police involved. Have a nice day!"

....

"Hello Sheriff? Angus here on Mongoose road. Those robot nutters have been up all night chanting 'Death To Humans' again. It's scaring my cows and putting off their milk production. Can you get those idiots to keep it down?"

....

"Neighbor, I can respect that you have problems with the government. I don't like 'em much either. But the next time I catch one of your super-intelligent mechanoids trespassing on my side of the fence, it's mine. I'm going to catch it and put it to work. Or maybe melt it down and sell it for scrap. If they're really that smart, they will understand that's what the fence is for and stop being nuisances."

$\endgroup$
4
  • $\begingroup$ The robots aren't explicitly built by one person. Once the first one is made, he has instinctual instructions to reproduce (like a human, only it doesn't take 9 months). The only limit on how quickly they can replicate is their supply of suitable bodies. Once they replicate enough, the problem becomes too much for police to handle. The AIs build up to tens of thousands in about a month. $\endgroup$
    – Grizzly
    Commented May 6, 2019 at 6:59
  • $\begingroup$ The point is that many local folks with the power to disrupt or intervene become aware of the situation very, very rapidly. And, depending upon the belligerence of Dr C or the AIs, some friendly police might drop by for a visit sooner than might be expected. $\endgroup$
    – user535733
    Commented May 6, 2019 at 13:45
  • $\begingroup$ They become aware of what's happening pretty quickly but these AIs are actually pretty cooperative at first. It's not such a pressing issue since the city isn't actually being invaded until Dr. C decides to go wild. Everyone spends about two weeks wondering what the AIs actually want and why they're not violent. One of the biggest parts of the novel is the reason why the US don't opt for a forceful, violent solution almost immediately, but it's a lot of extra explanation. $\endgroup$
    – Grizzly
    Commented May 7, 2019 at 14:08
  • $\begingroup$ Well, if the humans involved with the project get zoning variances and building permits and do the right things, then everyone in local government (and then media, and then neighbors) will know before ground is broken for the first building. If the humans try to hide the project (doing it right is expensive!), then local government will know before the first building is complete...and will be much more nosy about it. Also, the kind of fabrication under discussion is neither small nor cheap - the humans will need a lot of money, and investors can be be even nosier than local government. $\endgroup$
    – user535733
    Commented May 7, 2019 at 17:50
0
$\begingroup$

The details of how they would threaten or attack the nation should provide an answer.

If they plan on attacking electronically, whatever department with the most resources to combat cybercrime would be a good first guess.

If they somehow take over military hardware, then a more military-based response would make more sense.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .